search

SKKU-ESLAB / Auto-Compression

Automatic DNN compression tool with various model compression and neural architecture search techniques
MIT License
20 stars 20 forks source link

Add the resume option #14

Closed GH-Jo closed 2 years ago

GH-Jo commented 3 years ago

Training speed is slow because there are many candidates for each layer.

It is recommended to enable to stop the training at any epoch and resume training again.